Date: Sat, 5 Dec 1998 11:26:45 -0500
From: "Jay R. Ashworth" <jra@scfn.thpl.lib.fl.us>
To: "'nanog@merit.edu'" <nanog@merit.edu>
Subject: Re: More Sidgemore on per-bit pricing
[...]
Well, we'll see. The problem is suits who don't understand that
billing by the byte, rather than by the router port, usually has higher
costs than it does revenue enhancement. If I assert that I'm trying to
reduce customer's billings, then, as a CEO of a publicly held company,
I'd better have something up my sleeve that will actually enhance
revenue, or I'll be out of a job, real quick.
I suspect that "it's too hard and/or expensive to bill by byte (or
kilo-byte or mega-byte)" may become one of the great myths of the
Internet.
It should be noted that phone companies, particularly LECs, are probably
better than anybody at billing in very fine usage-basee increments.
"Suites" often understand costs and pricing very well. Suites often
also understand competitive advantage. I wonder if John Sidgemore is
thinking that offering fine-grained usage-based billing is something that
he can do well (if he leverages the billing skills of other parts of
WorldComm) and that his competitors won't be able to replicate, (e.g.,
the thousands of ISPs who seem to think that sending out one fixed-rate
bill a month is a sizable burden).
Beyond suspecting that fine-grained usage-based billing in the Internet
will be introduced by those that can do it well, I haven't a clue what
the long term future is...
-tjs
I agree 100%, it's one of the great "truisms" which has been repeated
over and over since the early '80's at least on lists like these.
In my experience it rests on largely a moralistic view rather than an
economic model. For example, the underlying presumption is that it's
somehow "wrong" to charge for the cost of billing (why?), and worse
yet to charge cost+profit on just the billing activity (why?) Yet in
essence every business which bills customers sells billing services at
a profit or they're not in business very long, if you want to look at
it like that.
I can certainly understand why someone would like to pay only for the
actual service or product and not pay for being billed, who wouldn't?
But if service+billing, where the billing is a larger percentage than
some other billing model, still is competitive (e.g. because it
reduces the cost of the service even more) then it's a potential
winner.
How the underlying charge pie-charts out in terms of cost-factors is
really of little concern to the customer if their final deal is
better. But that's basically what this "billing costs too much"
argument is often trying to say, that increasing the pie-chart slice
of billing is unreasonable, independent of how that affects the other
slices or the total cost (of course if the rest stays the same then
yes, it's probably a loser.)
Not at all, Barry. My assertion rests on two things:
1) Routers are too damned busy as it is; too busy, we're told, to run
the filters that would keep much of the crap off the net. It's
unlikely the money made by packing more customers into a given
amount of uplink would outweigh the costs of gathering and
processing the information at that fine a granularity.
2) The telcos currently control the local loop, and are pricing
that on a flat rate basis, mostly, frame and ATM
notwithstanding (there's _still_ a flat cost, somewhere).
I don't at all object to "usage-sensitive" pricing, burstable T's and
the like; I'm looking at one right now. It's this "slap a byto-meter
on it" mentality that demonstrated, I feel, a fundamental
misunderstanding of the net. But then, I expect that from telco suits.
Cheers,
-- jra